You are on page 1of 20

NASA Technical Memorandum 4720

A Historical Overview of
Flight Flutter Testing

Michael W. Kehoe

October 1995

NASA Technical Memorandum 4720

A Historical Overview of
Flight Flutter Testing

Michael W. Kehoe
Dryden Flight Research Center
Edwards, California

National Aeronautics and
Space Administration
Office of Management
Scientific and Technical
Information Program
1995

In the early years of aviation. elastic. Early test engineers were faced with inadequate instrumentation. excitation methods. High-speed aircraft are most susceptible to flutter although flutter has occurred at speeds of 55 mph on home-built aircraft. FLUTTER HISTORY The first recorded flutter incident was on a Handley Page O/400 twin engine biplane bomber in 1916. higher airspeeds and a shift from external wire-braced biplanes to aircraft with cantilevered . integrated effectiveness of the various approaches used. one still must fly close to actual flutter speeds before imminent instabilities can be detected. Aeroelasticity plays a significant role in the design of aircraft. The elevators on this airplane were independently actuated. The first formal flutter test was carried out by Von Schlippe in 1935 in Germany (ref. Wing-aileron flutter was widely encountered during this time (ref. Control surface flutter began to appear during World War I. First. and T-tail configurations increases the likelihood of flutter occurring within the desired flight envelope. instrumentation systems. A rise in amplitude would suggest reduced damping with flutter occurring at the asymptote of theoretically infinite amplitude as shown in figure 1. Maximum response amplitude 0 Vflutter Airspeed INTRODUCTION Aeroelastic flutter involves the unfavorable interaction of aerodynamic. considerable improvements have been made in flight flutter test technique. no formal flutter testing of full-scale aircraft was carried out. Although some mild instances of control surface flutter were encountered afterward. Flutter testing. 1). instrumentation systems. These analytical results are often verified by wind-tunnel flutter models and ground vibration tests. Third.SUMMARY This paper reviews the test techniques developed over the last several decades for flight flutter testing of aircraft. Von Baumhauer and Koning suggested the use of a mass balance about the control surface hinge line as a means of avoiding this type of flutter. these were usually eliminated by increasing the mass balance of the control surface. and parameter identification algorithms (for frequency and damping estimates from the response data) are described. and inertia forces on structures to produce an unstable oscillation that often results in structural failure. Flight flutter testing provides the final verification of the analytical predictions throughout the flight envelope. Structural excitation systems. This idea was applied successfully to several German aircraft until a Junkers JU90 fluttered and crashed during flight tests in 1938. The introduction of thinner wings. The flutter mechanism consisted of a coupling of the fuselage torsion mode with an antisymmetric elevator rotation mode. Von Schlippe’s flight flutter test method. 3). all-movable horizontal and vertical stabilizers. 2). including a history of aircraft flutter incidents. Practical experiences and example test programs illustrate the combined. comments regarding the direction of future developments and needs are presented. is still a hazardous test for several reasons. Second. The solution to the problem was to interconnect the elevators with a torque tube (ref. In fact. and stability determination methods is reviewed as it pertains to flight flutter testing. Today’s aircraft designs undergo sophisticated aeroelastic analyses to ensure that the design is free of flutter within the flight envelope. instrumentation. Since then. The development of excitation systems. the aeroelastic stability may change abruptly from a stable condition to one that is unstable with only a few knots’ change in airspeed. After World War I. Finally. and stability determination techniques. The aircraft was simply flown to its maximum speed to demonstrate the aeroelastic stability of the vehicle. His approach was to vibrate the aircraft at resonant frequencies at progressively higher speeds and plot amplitude as a function of airspeed. This paper presents a historical overview of the development of flight flutter testing. no speed regime is truly immune from flutter. digital data preprocessing. 940475 Figure 1. and response data analysis. however. subcritical damping trends cannot be accurately extrapolated to predict stability at higher airspeeds.

Von Schlippe’s technique consisted of exciting the structure using a rotating unbalance weight. This type of instability could lead to abrupt fatigue failure. DEVELOPMENT OF FLIGHT FLUTTER TEST TECHNIQUES Another form of flutter dealt with in the 1930’s was servo tab flutter. Seven incidents of flutter from 1947 to 1956 involved the carriage of external stores. Von Schlippe conducted the first formal flight flutter test in Germany in 1935. The forced response amplitude would rapidly increase as the aircraft approached its flutter speed. on primary lifting surfaces as for the F-117 stealth fighter (ref. 6). while flight testing the new P-80 airplane. AT-8 flight test Wright Field. 4). however. shows the response amplitude data as a function of airspeed. This prediction was correct. the T-46A trainer experienced aileron flutter during a test flight that was being flown to find the proper amount of mass balance. F-18. and F-111 aircraft produce an aeroelastic instability known as a limit cycle oscillation (LCO) (refs. 5). 5). 2 . both of which experienced vertical fin flutter. Inadequate structural excitation equipment and unsatisfactory response measurement and recording equipment were identified as probable causes for this accident (ref. Air racers experienced many incidents of flutter from the mid-1920’s until the mid-1930’s as attempts were made to break speed records. and then recording the response amplitude as a function of airspeed. as well as pylon-mounted engines (ref. Although these oscillations are mostly characterized by sinusoidal oscillations of limited amplitude. taken from reference 4. Even today servo tab flutter is still a problem. 3) predicted that this type of flutter would be around for many years. Flight flutter incidences Figure 2. Nov. Figure 2. Supersonic flutter then began to be studied more seriously as these speeds became routinely flown. mph 320 940476 Much has been learned about the prevention of flutter through proper aircraft design. Certain combinations of external stores carried by the F-16. v. 11) aircraft. In 1986. The Germans successfully used this technique until 1938 when a Junkers JU90 aircraft fluttered in flight and crashed. Today. 8 and 9). 7). NACA pilots reported an incident of aileron buzz (ref. In the 1950’s a fighter airplane was lost because of a failed hydraulic line that was attached to a panel that had experienced such panel flutter (ref. still occur. and 5 give many examples of these incidents. for between 1947 and 1956. flight testing has shown that the amplitudes may either decrease or increase as a function of load factor (angle of attack) and airspeed. measuring the response amplitude. The graph shows that destructive flutter for this airplane was averted by the narrowest of margins during this flight test. Therefore. there were 21 incidences of flutter involving transonic control surface buzz. Supersonic speeds also produced a new type of flutter known as panel flutter. From 1947 to 1956. In 1944. 1941 x Left wing data Right wing data 50 x 40 30 Amplitude at v Amplitude at 100 mph 20 10 0 x x x x x 80 160 240 Velocity. 4). 4). Response amplitude ratio as a function of airspeed (ref. The stores carriage problem is still significant today. References 3. the flutter speed could be estimated from data obtained at subcritical airspeeds. These ailerons were free floating and driven by tabs at the trailing edge of the aileron (ref. Panel flutter involves constant amplitude standing or traveling waves in aircraft skin coverings. 5). 4). so the avoidance of panel flutter is important.wings resulted in more wing flutter incidents. The usual practice at this time was to fly the airplane to the maximum speed and then to observe the stability of the structure. The objective of his test method was to lessen the risk associated with flutter testing. The carriage of external stores affects the aeroelastic stability of an aircraft. Primary surface flutter began to appear around 1925 (ref. Collar (ref. 10) and the E-6 Tacamo (ref. Chuck Yeager first achieved supersonic speeds in level flight in 1947. particularly with the many store configurations that an airplane can carry. the transonic flight regime is still considered the most critical from a flutter standpoint. 4. The United States attempted this technique in the 1940’s with flutter tests of a Martin XPBM-1 flying boat and a Cessna AT-8 airplane (ref. New aeroelastic problems emerged as aircraft could fly at transonic speeds. 11 cases of tab flutter incidences were reported for military aircraft alone (ref. Prototypes of both the F-100 and F-14 fighters had incidences of rudder buzz.

The use of oscillating vanes to excite the structure was being considered during this time. The P6M aircraft program took a departure from this methodology (ref. excitation systems consisted of inertia shakers. Frequency sweeps were made to identify resonances. Frequency and damping are The F-111 program is an example of this procedure. These sweeps were often followed by a frequency dwell–quick stop at each resonant frequency. many aircraft were equipped with excitation systems. Instrumentation had improved and response signals were then being telemetered to the ground for display and analysis. Some programs still displayed response signals on oscillographs in the airplane. This flight flutter program used random atmospheric turbulence to excite the structure and spectral analysis to analyze the response data for stability. Many experimenters realized the importance of adequate structural excitation for obtaining a high signal-to-noise ratio (ref. 14). 13). Filtered and unfiltered accelerometer response data were displayed on strip charts and on X-Y frequency sweep plotters. taken from Onboard recorder and TM transmitter Aero tab excitation Ground control and data processing TM receiving and tape recording Test director Band pass filters X-Y frequency sweep plotters Band pass filtered to accel choice Record two at one time Brush recorders 16 channels accel and aero tabs 8 channels flight parameters and control system (unfiltered) Postflight Airplane tape Brush record of all transducers Manual data reduction (g versus Mach) 940477 Figure 3. Computers were not used for analysis of the data. In-flight analysis was usually limited to log decrement analysis of accelerometer decay traces on strip charts to determine damping. Damping was manually determined from the frequency dwell–quick stop decay traces. 3 . manual control surface pulses. Since the 1970’s. The computer has allowed for the rapid calculation of the fast Fourier transform (FFT). Computers have fostered the development of more sophisticated data processing algorithms that are useful for analysis of response data from either steadystate or transient excitation. reference 13. The objective of this technique was to use every minute of flight time for dynamics data and to eliminate special test points for flutter. From the 1950’s until the 1970’s. F-111 flight flutter test procedure (ref.In the late 1950’s. and thrusters (bonkers). 12). Figure 3 shows a schematic of the process. digital computers have significantly affected flight flutter testing techniques.

6 1.4 1. Detection of impending aeroelastic instabilities cannot be made without adequate excitation.8 Mach number 940478 Figure 4. Technology development associated with each component will be reviewed and a discussion of the impact on the safety of flight flutter testing will follow. Hz Frequency. 8 Hz 4 50 0 . have increased the total time to clear the flight flutter envelope. however. Figure 4 shows this process. the time required to conduct flight flutter testing per test point has decreased. EXCITATION SYSTEMS Structural excitation is a necessary part of the flight flutter testing methodology. 15). and real-time frequency analyzers. data are telemetered from the airplane and simultaneously displayed on strip charts. The test director.8 1. Such analysis is done online in a near-real-time manner. response measurement. In this figure. Although flight flutter test techniques have advanced. The test technique typically used is to monitor the telemetered response signals with real-time frequency analyzers and strip charts.now estimated with parameter identification techniques. 4 . These data are also acquired by digital computers that process the data using parameter identification techniques for estimating frequency and damping (ref. The ability to analyze more data at each test point and the increased number of flight test points resulting from more sophisticated aircraft designs. Frequency and damping trends are established as a function of airspeed or Mach number.0 1. who communicates with the test aircraft. and data analysis for stability. Typical modern flight flutter test process.04 Right plus left wingtip Strip charts 0 12 Right minus left wingtip 0 25 Frequency. These trends are extrapolated to determine the stability at the next higher airspeed test point.12 Data Damping. .2 1. has access to all of this information to make decisions on continuing the flutter envelope expansion. today’s techniques are still based upon the same three components as Von Schlippe’s method: structural excitation. As computer speeds have increased. A computer acquires the data for analysis to determine frequency and damping estimates. Adequate excitation provides energy to excite Data Data Real-time frequency analyzers Right wingtip Spectral analysis facility ground station Stability trends .08 g .

Flight control surface pulses are still used today as excitation for flight testing. Such pulses approximate a delta function that theoretically has a high frequency content. The XF3H-1 airplane used an oscillating rudder for excitation to investigate a rudder buzz instability (ref. the A-7A (1965). such pulses often do not provide an adequate level of excitation to determine the onset of flutter. The F-101 (mid-1950’s). The thrusters excited flutter at a speed 5 knots below that where the structure was previously excited by rudder pulses without incident. however. In a subsequent flight. so many can be applied at each test point. control surface pulses have continued in use to excite the structures of many airplanes since 1950. It is difficult for any one system to meet these requirements simultaneously. the F-16XL airplane flight control system has a 1. and thus the degree of excitation is inconsistent. The structure was excited using rudder pulses and by thrusters (impulse generator). Over the years. inertial exciters. some of the early Boeing 747 flutter testing (1969). 16) clearly demonstrated the importance of adequate excitation levels in 1986. and (3) have power requirements (electric or hydraulic) that the airplane can meet. The fact that flight was possible beyond the flutter speed without exciting flutter with pulses was demonstrated during a flight flutter test in the 1950’s (ref.6-Hz low-pass filter (single pole) in the stick input path that would washout any sharp stick motion commands to the control surface actuators. 19). Manual control surface pulses are still used today on most small aircraft and sailplanes because this is usually the only affordable type of excitation for these aircraft. During flutter testing of the B-58 airplane (ref. This airplane was excited on the initial flight tests by control surface pulses and random atmospheric turbulence. 4). These weather conditions provided higher levels of excitation than the levels induced during the flight flutter test. In addition. In the 1930’s. In the mid-1960’s. These function generators provided signals to oscillate the horizontal stabilator and the ailerons of the F-4 airplane (ref. For example. aerodynamic vanes. The rudder was oscillated by supplying a variable frequency command signal into the rudder servo of the autopilot system. many types of excitation have been tried with varying degrees of success. Test duration for each pulse is short. it is difficult to get repeatable pulses. (2) be lightweight so as not to affect the modal characteristics of the airplane. the estimated values suggested lower aerodynamic damping than actually existed. either the pilot cannot provide a sharp enough input or the control system is unable to provide a sharp enough disturbance to excite any critical flutter modes above 10 Hz. all of the selected vibration modes with sufficient magnitudes to accurately assess stability from the response data. oscillating control surfaces. it was found that a structural excitation level at least three to four times higher than was obtained by random atmospheric turbulence was necessary to provide an acceptable level of excitation. 17). the Germans decided that improper exciter location resulted in poor responses that prevented the determination of the onset of flutter and sometimes resulted in flutter occurring unexpectedly (ref. Flutter did not occur during the flight test. The Transavia PL12/T-400 airplane (ref. and random atmospheric turbulence. The stabilator could 5 . the early testing of the F-4 aircraft (late 1950’s). Some more common means include control surface pulses. This system could excite over a frequency range of 5 to 35 Hz and with the frequency stepped every 3 sec by an automatic rotary switch located in the cockpit. In the 1950’s. Second. the United States learned that low excitation levels tend to give a large scatter in the damping values estimated from the response data. have lowpass filters in the stick input path that filter out highfrequency signal content. and low-speed testing of the DC-10 airplane (1970) all used control surface pulses for structural excitation. 18). the airplane experienced violent oscillations of the rudder and tail boom when it was flown in rough weather conditions. Control Surface Pulses Manual control surface pulses were the first means of excitation. thrusters. Two benefits of this type of excitation are that no special excitation equipment is required and that the transient response signature of the structure is easy to analyze for stability. First. The purpose of this particular flight was to investigate the stability of a vertical stabilizer. This provided sudden control surface movements. 19).There are several drawbacks. Third. Oscillating Control Surfaces Commanded oscillations of the control surfaces were also used in the 1950’s. however. In spite of their limitations. modes up to about 10 Hz can be excited this way. Most modern fly-by-wire flight control systems (analog and digital). The excitation system must not only provide adequate force levels but must also (1) provide adequate excitation over the desired frequency range of interest. Depending on the type of control system. electronic function generators were developed to provide signals to control surface servos in the autopilot system.

thrusters have not been used in the United States for any major flight flutter test program. 4). As a result. Thrusters were also used for portions of the F-4 flutter testing in the early 1960’s. Use of these devices was partially successful for this program (ref. lightweight devices that generally do not affect the modal characteristics of the airplane. sometimes known as bonkers. F-15. A cockpit controller provided the capability to adjust the excitation amplitude. 19) on several airplanes to investigate flutter characteristics. the excitation capability may be limited at lower frequencies and excessive at higher frequencies. In addition. The magnitude of the forces required to adequately excite an airplane is usually very large. three on top and three on the bottom. particularly in exciting the higher frequency modes of the airplane.5 in. 19). 6 . the flutter speed of the airplane is not affected as it might be with other types of excitation systems. The overall dimensions were 4. Thrusters were used for part of the flutter clearance of the F-101 airplane in the mid 1950’s. are an early device circa (1940) used for structural excitation. The F-18 aircraft excitation system can generate sinusoidal and bandpass-filtered pseudo-random commands to the flight control surfaces (ref. This method of excitation has been successfully used for the X-31 and YF-22 airplanes and for the stores clearance work on the F-16. For example. the combination of aerodynamic force at low frequencies and control surface inertia forces at higher frequencies provided adequate excitation for this airplane (ref. As a result.. as was the case for the F-4 flutter testing (ref. Thrusters Thrusters. The precise control of frequency was difficult with the equipment available.5 by 4. These shakers were hydraulically powered and electrically controlled and were used to excite a frequency range of 5 to 40 Hz. the rotating unbalance equipment designed for (but never installed on) the XB-36 had a maximum force output of 1000 lb. Further advances in electronics during the 1970’s and 1980’s resulted in the ability to send excitation signals to the control surface actuators that were not sinusoidal.000 lbs (ref. The disadvantages for these devices include single-shot operation. The actuators for each surface were modified to provide the required gain and frequency response. The force output was 40 lb at 7. the hardware required to produce the unbalance forces often could not be contained within the wing contour. which is important when the airplane has to dive to attain a test condition. 20). The inertia force is proportional to the exciter weight multiplied by the square of the rotating speed. 21). As a result. Inertia shakers were used for the B-58 flutter testing in the early 1950’s. These small. Thrusters are simple. 19). and installation in each wing weighed between 400 and 500 lb (ref. 18). These devices produce transient responses of short duration. and dwell frequency. The rotating unbalance exciter was widely used for flight flutter testing in the 1940’s and 1950’s. these systems often are very heavy and raise concerns about the effect on the modal characteristics of the airplane that they are installed on. was by observing a meter in the cockpit that measured the response of selected vibration pickups during flutter tests (ref. ballistic exciters or impulse generators.5 Hz and 150 lb at 40 Hz.sweep from 8 to 30 Hz and the ailerons from 2 to 16 Hz. Often. special actuators are required to excite critical highfrequency modes.1 to 100 Hz. start and stop sweep frequency. Although significant actuator roll-off occurred above 20 Hz. 12). The Martin XPBM-1 flying boat flight flutter test program used such a system. solid-propellant rockets have burn times of 18–26 msec and maximum thrust levels of 400–4. One method of tuning the exciter frequency. A disadvantage of this type of system is the frequency response limitations of the control surface actuators. These systems derive their forces from mass reactions. the force level increased linearly between these two frequency values (ref. The primary advantage of this type of system is that no additional hardware is required except for an excitation control box located in the cockpit. Since then. thus the exciter would not stay tuned on the resonant frequency. and each unit weighed 25 lb. and F-18 airplanes. and their inability to provide a wide frequency band of excitation. Six thrusters were mounted on each wingtip. Inertial Exciters Thrusters were also used by Douglas Aircraft Company in the 1950’s (ref. The X-31 airplane could sweep frequencies from 0. single-shot. mode selection (sweep or dwell). A large variety of rotating eccentric weight and oscillating weight inertia exciters have also been tried. This type of shaker worked well.5 by 8. difficulty in firing two or more either in phase or out of phase with respect to each other. Usually required are thrusters with three different burn times to excite modes in a frequency range that covers 5 to 50 Hz. as done by a Convair F-92 pilot in 1950. 17).

24). the F-111 horizontal and vertical stabilizer. electronically controlled. The excitation frequency and amplitude at a given airspeed can be controlled. 26). 6) was developed by W. the disturbance of the normal airflow around the wingtip or stabilizer tip with the vane present.4 to 10 Hz. The weight of the shaker was 8. oscillating mass exciters.The Convair F-102A. The advantage of this type of system is that it can excite low frequencies well. show the characteristics of some of these excitation systems. Reed (ref. and oscillates about some mean angle. In the frequency range between 5 and 50 Hz. Boeing 747. many flight flutter test programs have used the aerodynamic vane as a means of excitation. This wingtip unit weighed 150 lb and was mounted on the right wingtip only. This system was used for the flutter clearance of a F-16XL airplane with a modified laminar flow glove (ref. After the 1950’s. Aerodynamic vanes were first used in the 1950’s. and the X-29 flaperon. Oscillation of the vane will result in a varying aerodynamic force acting on the airplane. Since then. which was flutter tested in the late 1950’s. inertia exciters were not used extensively. The YB-52 airplane used a wingtip oscillating airfoil shaker for flight flutter testing (ref. Inertial excitation system (ref. One exciter was placed at each wingtip. The amount of force depends on the size of the vane.5 in. The vane/cylinder assembly weighs approximately 10 lb. 7 . This system produced periodic excitation to the structure and was successfully used for flight flutter testing. 13). one at each horizontal stabilizer tip. which has a depth of 4. and the force time history produced is repeatable. the amplitude at high frequencies is limited only by the response characteristics of the vane drive mechanism. driven either electrically or hydraulically. 23). Typical sweep times were approximately 7 min. There are two notable variations of the oscillating aerodynamic vane concept: a rotating vane and a fixed vane with a rotating slotted cylinder attached to the vane trailing edge. This shaker was sufficiently compact to be installed inside the wing. and T-46A. L-1011. The excitation frequency could be varied from 1. Hydraulically actuated movable mass (wand) 940479 Figure 5. and both sets of vanes were synchronized to provide either symmetric or antisymmetric excitation. F-111. Each wingtip and horizontal stabilizer tip could be operated in and out of phase with respect to each other. However. The main disadvantage is that the maximum force produced varies with the square of the equivalent airspeed. Aerodynamic Vanes An aerodynamic vane consists of a small airfoil that is usually mounted to the tip of a wing or stabilizer. The B-1A airplane also used an inertia shaker system (fig. there has been limited use of such systems to provide partial structural excitation for the F-14 horizontal stabilizer. also used inertial shakers. The device generates periodic lifting forces by alternately deflecting the airflow upward and downward through the slot. 25). dynamic pressure. A-10. A similar amount of weight was installed on the opposite wingtip. Other disadvantages include the addition of mass. 5) for flight flutter testing. Frequency sweeps were conducted from 5 to 35 Hz for this program. 22). and the large power requirements usually needed to operate this system. This system was capable of producing a maximum force of about 550 lb of force with an exciter weighing approximately 40 lb (ref. The system uses exceptionally little electric power and is easy to install. Boeing 757.tip. The vanes were continuously rotated through 360 degrees. The system consisted of five hydraulically driven. The vane is generally mounted on a shaft. This system adequately excited the modes of interest and was essential for safely expanding the flight envelope. F-14. C-17.5 lb (ref. The fixed vane with a slotted rotating cylinder attached to the trailing edge (fig. The C-5A airplane excitation system consisted of a rotating vane mounted on top of the wing at each tip and on top of each horizontal stabilizer at each tip (ref. S-3A. the force varied from approximately 20 to 300 lb. 23). These programs include the DC-10. and one at the tip of the vertical stabilizer. Tables 1 and 2 which were taken from reference 13. and angle of rotation.

Turbulence excites all of the surfaces simultaneously. Summary of aerodynamic excitation systems used in U.Table 1. This approach was tried in the late 1950’s. which causes both symmetric and antisymmetric modes to be excited at the same time.S. vertical tail surfaces Inboard on stabilizer near side of fuselage. Atmospheric turbulence has been used for structural excitation in many flight flutter test programs (ref. Turbulence is generally produced by 8 . Objectives of this technique were to use every minute of flight time for dynamics data and to eliminate special test flights for flutter.5–18 Hz 3–25 Hz 90 Linear period C-5A Wing stabilizer External vanes on top of surfaces near tips .5–7. sec Sweep law Aero-tab Table 2. the extent of turbulence within an air mass can vary widely.0 Hz 90 Exponential 1–20 Hz and 1–10 Hz 90 Exponential 90 Wing stabilizer External vanes 1–18 Hz 3–25 Hz 90 30 Linear period S-3A Side of fuselage under stabilizer External vanes 1. 15). The greatest attraction to this type of excitation is that no special onboard exciter hardware is required.S. Reference 27 provides an excellent description of turbulence and its use for excitation in flight flutter testing. Airplane Surface Location 747 Wings External vanes at wingtips Wings horizontal Vertical tail External vanes at tips of main surfaces DC-10 L-1011 Frequency range Time to sweep. The P6M Seamaster flight flutter program (ref. 14) used random atmospheric turbulence to excite the structure and spectral analysis to analyze the response data for stability.5–25 Hz 60 normal 30 dive only Exponential F-14 Wing fin Aero-tab External vane 5–50 Hz 15 Exponential F-15 Normal control Ailerons Stabilator 2–16 Hz 5–10 Hz 100–200 45 Linear frequency F-111 Wing 35–2 Hz 45 Exponential Frequency Range Time to sweep. top of fin 35–2 Hz 45 Exponential Random Atmospheric Turbulence weather-front winds and thermal activities. Airplane Surface Location F-14 Horizontal tail Right side stabilizer only 5–50 Hz 15 Exponential F-111 Horizontal. Summary of inertial excitation systems used in U. sec Sweep law 1. The YF-16 also used random atmospheric Natural turbulence is the random variation in wind speed and direction. This method eliminates the need to perform symmetric or antisymmetric sweeps.

Long data records are required to obtain results with a sufficiently high statistical confidence level. F-16XL with vane and rotating slotted cylinder excitation system (ref. A comparison of response data damping values from random atmospheric turbulence and vane excitation (fig. there are several disadvantages. Turbulence usually excites only the lower frequency modes for most airplanes. The response data must be measured at enough locations and be of high enough quality that the flight can be conducted safely. excitation levels usually give a large scatter of the estimated damping values. and turbulence often interferes with other engineering disciplines data. which makes data analysis very difficult. Included in the instrumentation system is the measurement. telemetry. caution should be used when using this form of excitation to ensure that the critical flutter mode has been excited throughout the flight envelope. and F-15 short takeoff and landing/maneuver technology demonstrator) have been cleared using atmospheric turbulence. recording. This flight test program also illustrates the lessons learned during the B-58 flutter test program. and displaying of the flight data. X-29A. 8) shows that the turbulence response data.Figure 6. while only the 8-Hz mode was well excited by turbulence.e. Although this method has been used with some success over the years. Inadequate 9 .. and the estimated damping values often indicate lower damping than actually exists. which have a lower excitation level. 26). This data comparison clearly indicates the poor data quality that is typically obtained with random atmospheric turbulence. The turbulence was reported to be light-to-moderate for this flight condition. INSTRUMENTATION The instrumentation used to record the structural responses of an airplane to excitation is another critical component of flight flutter testing methodology. Schweizer SA 2-37A motor glider. The signal-to-noise ratio of the response data is often low. All of the structural modes were excited by the vane. 28). Figure 7 compares the power spectrum plots obtained from the F-16XL airplane excited by random atmospheric turbulence and by a vane with a slotted rotating cylinder attached to the trailing edge. The turbulence that is found is often not intense enough to produce sufficient excitation compared with that obtained with onboard exciters. Flight time is lost looking for sufficient turbulence. advanced fighter technology integration (AFTI) F-111 mission adaptive wing. turbulence along with the random decrement technique for data analysis to clear the flutter envelope of the basic airplane (ref. AFTI F-16. Although the flutter envelopes of many modern aircraft (i. consistently have lower estimated damping values.

9 1. 26). wide).0 Mach number 1. 26).6 1. To minimize the effects of the outside air temperature on the damping. . Measurement The calibration of these devices drifted during operation mainly because of the electronics in use then. As an example.7 . Subminiature accelerometers were developed later and these were lighter but still were large (1/2-in. high. high. have high sensitivity.03 . and weighed about 1 lb. .07 . the accelerometers were about 3 in. these units had built-in electric heaters to maintain a constant temperature of 165° F (ref.000 Hz.02 . .1 Atmospheric turbulence excitation Forced excitation using linear sine sweeps .1 1.8 . Piezoelectric accelerometers have been developed such that miniature units today weigh less than one-tenth of an ounce.9 and 30.05 Structural . have a linear frequency range of 1 to 10.9 1.7 .01 0 .02 g2 . (b) Random turbulence.03 Amplitude. Hz 60 920769 920768 (a) Exciter sweep..03 .0 Mach number 1.01 0 . 10 . Accelerometers used in the early 1940’s were large and heavy. These were fluid damped devices.01 0 10 20 30 40 50 Frequency.6 .02 g2 Amplitude. The selection of the device to use often depends on the ease with which installation can be accomplished.8 .04 damping . 2 in.04 .04 damping .07 .08 .05 Structural .2 920771 (b) Antisymmetric.2 920770 (a) Symmetric. Figure 7. The accelerometers used for the B-58 flutter testing in the 1950’s were of the strain-gage type.02 . 17).03 . wide.000 ft (ref. Structural damping values for wing bending modes (ref.06 . Comparison of excitation of the left-wing aft accelerometer at Mach 0.04 . Figure 8.06 . and have amplitude linearity from 1 to 500 g. Hz 0 60 10 20 30 40 50 Frequency.08 .01 . operate in a temperature range of –65 to 200° F. and 1-in. Atmospheric turbulence excitation Forced excitation using linear sine sweeps . The most commonly used transducers to measure the excited response of a structure have been the accelerometer and strain-gage bridge. although today the more commonly used device is the accelerometer.

4) in that decade. 30. Sometimes the pilot had a small cathode ray tube oscilloscope in the cockpit to display the decay trace of a single. 30) and then to display these signals on strip charts for the flutter engineer. Today the available portable digital recorders are compact and can acquire data from flight instrumentation for storage within the unit’s computer memory. processor. airspeed. subtract. and filter the data signals telemetered (ref. computer technology had advanced to the point that it became feasible to use a computer to perform online stability analysis of the flight flutter test data.25 in.15 in. which mechanically actuated a recorder installed in the observer’s seat of the airplane (ref. data were commonly FM/FM telemetered to the ground. Pulse code modulation (PCM) or digital telemetry was initiated in the 1960’s. wide and 0. Strip chart capabilities have greatly increased. the accelerometer responses were recorded on photographic oscillographs mounted in the cockpit or airplane cabin. PCM telemetry is now usually preferred for most flight flutter testing. and then displayed on strip chart recorders. Computers were used to manipulate and display data to some extent during the 1950’s. The pilot then uses a cockpit display as an aid to reach and hold desired test conditions. and 32). In the 1970’s. and today strip charts continue to be the primary device to display real-time accelerometer and strain-gage response time histories. Today’s accelerometers accurately measure the structural response in almost any kind of environment. Telemetry and Recording Displays Recording equipment was not adequate during the 1930’s and was cited as a possible reason that the Germans lost several airplanes during flutter testing (ref. it became possible to provide the pilot a real-time guidance system for maintaining flight conditions. Many basic airplane parameters. In the 1950’s. In the 1950’s. 31. These devices required a developing time for the paper. In the 1980’s. could now be displayed on cathode-ray tubes. With this system the pilot flies the airplane to minimize the computed differences between the desired and actual flight condition (Mach number and altitude). During the 1940’s. such as Mach number.Typically the dimensions are as small as 0. Advances such as this will significantly reduce the cost associated with flutter testing. and fuel quantities. The recording system used was a thin wire attached to the rudder. In the 1930’s. The computed differences are telemetered to the airplane from a ground-based computer. The frequency bandwidth of PCM systems had increased significantly by the 1980’s. high. they have been used for low-cost flutter testing (for example. As a result. Analog computers were sometimes used to add. although FM/FM telemetry was still widely used for flutter testing because of the frequency bandwidth required. sensor. Each unit contains a battery. 29). Although these units can acquire a limited amount of data. integrate. The ground tapes were typically noisier than the onboard tapes mainly because of datatransmission problems. angles of attack and sideslip. The telemetry systems during this period were small and typically only 8 to 12 channels of data could be telemetered to the ground (refs. The data from the unit’s computer memory are downloaded after each flight into a digital computer for analysis. These devices do not require any wiring and the signals may be transmitted either to a receiver located on board the airplane or directly to a receiver located on the ground within a small flight radius of the transmitter. a Junker JU86 airplane underwent flight flutter testing to identify the effects of balance weights on rudder stability. Subminiature instrumentation has also been developed as self-contained peel and stick devices (ref. the Pond Racer airplane). A frequency 11 . on-board tape recorders captured all of the flight data while the recorders on the ground captured only the data that could be sent down from the airplane. Computers were also used to provide discretes and alphanumerics in the control room. data were primarily displayed on strip charts in the control room for analysis by the flutter engineer. multiply. 33). This display provided the flutter engineer with the ability to more closely monitor the flight test conditions of the airplane. bandwidth of 200 Hz is easily attainable and sufficient for most flutter applications. Direct storage of the data eliminates the need for expensive equipment associated with PCM systems. As a result. selected accelerometer (ref. the time history responses could not be viewed immediately as a result. The pilot was briefed before each flight by the flutter engineer on the anticipated response amplitudes and safe operating limits. 4). and transmitter. The PCM telemetry significantly increases the number of parameters that can be transmitted to the ground but requires a filter to prevent frequency aliasing of the analog response signal during digital sampling on board the airplane. antenna. recorded on magnetic tape.

The damping estimation consisted of using the log decrement method on the decay portion of a time history response. Evolution of instrumentation. and having ground stations with highly specialized test capabilities has significantly reduced the hazard of flight flutter testing. having increasingly reliable and accurate instrumentation. 22. but the analysis was generally conducted on the ground. because no computers with sophisticated identification algorithms were available. These data were recorded on an oscillograph recorder. The next component of flight flutter testing methodology is the analysis of the response signal. transient responses caused by either impulse input or exciter frequency dwell– quick stops. Figure 9 shows how aircraft instrumentation. discretes. and then clear the airplane to the next higher airspeed (refs. data recording and real-time support requirements. The accurate and timely evaluation of As telemetry became more available in the 1950’s. video Spectrum analyzers Analog stripcharts Digital computer to manipulate data FM telemetry X-Y plotters Fliters On-board aircraft tape 1960 Accelerometers and strain gages 1950 FM telemetry (10-20 channels) On-board aircraft tape Analog stripcharts Analog computer to manipulate data X-Y plotters Fliters Radio On-board photographic Flutter Accelerometers oscilloscope engineer on and strain gages <10 channels board the aircraft of data 1940 950442 Figure 9. video Spectrum analyzers Computer disk PCM telemetry (100s of channels) Alphanumerics. and real-time support in the ground station have changed since the 1940’s. All of this analysis was usually done by hand between flights. data recording. Occasionally. and 24). 17. 12 . DATA REDUCTION METHODS In the 1930’s and 1940’s. the data analysis methods just described were still used. The response signal can consist of random response caused by atmospheric turbulence or exciter input. 4). or steady-state responses caused by exciter frequency sweeps. removing the flutter engineer from the airplane during the test. Instrumentation Data recording Real-time support 1990 Digital stripcharts Interactive graphics Digital computer for data manipulation and analysis PCM telemetry (100s of channels) Accelerometers and strain gages FM telemetry On-board aircraft tape 1980 Analog stripcharts PCM telemetry (100s of channels) Accelerometers and strain gages Digital computer for data manipulation and analysis FM telemetry On-board aircraft tape 1970 Accelerometers and strain gages Year Alphanumerics. It became more common to excite the airplane at stabilized test points. discretes. In summary. this data to determine stability is critical to the overall safety of the flight flutter test program. the methods used consisted of measuring the response amplitude caused by a frequency sweep or determining the damping from a response caused by a control surface pulse. in lieu of telemetry. the flutter engineer was on board the airplane to do these analyses (ref. analyze the response for stability.this has resulted in exceptionally accurate stabilized flight test conditions (ref. 34).

reducing the modal density of the response signal (ref. KEAS 380 940484 Figure 10. however. An examination of several flight flutter test programs shows the effectiveness of the techniques used today to warn of the onset of flutter. Instead. Such a technique was used in the 1960’s (ref. during the flutter testing of the KC-135 airplane The F-14 and F-15 aircraft programs were among the first to take advantage of this advance in technology. The symmetric and antisymmetric modes could then be separated. also took advantage of the increased capabilities of the digital computer to increase the efficiency and safety of flight flutter testing. providing the capability to obtain frequency content of acquired signals in less that a second. engineers. no phase or damping information was available from this approach. proved to be difficult to separate for damping estimation using these analysis techniques. are limited to real-time monitoring of sensor responses on strip charts. 300 320 340 360 Airspeed. 28). the F-14 required 489 shaker sweeps to clear the basic airplane flight envelope. These manual or analog analysis techniques continued to be used until the 1970’s. 38). Figure 10 shows the frequency and damping trend information obtained. This practice is questionable because actual damping trends can be nonlinear. 13 . In the 1950’s. . 31) and spectral analysis (ref. response data are not being quantitatively analyzed. Modal damping was not estimated from the spectral analysis technique. Filtering of the response data was common to reduce the scatter in the damping estimates. Current State-of-the-Art Today. and 264 shaker sweeps were required for the Gulfstream II ER airplane (ref. vector plotting (ref. relying on intuition and experience. 32). For the F-14. The F-15 program used an analysis technique to predict the flutter boundary based on frequency and damping data acquired at subcritical speeds (ref. Data are analyzed at these points only. Autopilot off Autopilot on . During this phase. By 1970. The F-15 required 132 shaker sweeps and 156 frequency dwells to clear the basic airplane flight envelope (ref. Frequency and damping trends established from flight data. The data obtained at each stabilized test point establish a damping trend as a function of airspeed. The number of stabilized test points required to clear the flutter envelope of an airplane is typically high and consequently requires many flights to accomplish them. Even so. digital computer systems could be used for interactive analysis of flight data. Information is then extrapolated to predict the stability of the next planned test point.08 g . such as the YF-16 (ref. This analysis only provided the frequencies and amplitudes of the response signal being analyzed. 2.9 Hz 2. Closely spaced modes. 14) were also used to determine stability. Other programs.5 280 Since the 1970’s. Using tracking filters during an exciter frequency sweep improved the data quality obtained. 37). 41). Another method for enhancing the data analysis was to add and subtract time history signals. The most critical part of expanding the flutter envelope is the acceleration from one test point to the next. an equation error identification technique was used to estimate frequency and damping information (ref. During the early 1970’s.7 2.CONCLUDING REMARKS Data transmission problems usually added noise to the telemetered data. 177 shaker sweeps were required for the Gulfstream III. the typical approach to flight flutter testing is to fly the aircraft at several stabilized test points arranged in increasing order of dynamic pressure and Mach number.12 Damping. The speed of computers then allowed parameter identification algorithms to be programmed for estimation of damping from the response signals.04 0 3. 19) to filter data provided to an analyzer to produce a real-time plot of frequency and amplitude. For example. Data were often passed through a filter to reduce the data to a single-degree-of-freedom response. 37). in near-realtime. References 39 and 40 are excellent reports on modal parameter estimation and provide numerous references for the many different approaches taken. the fast Fourier transform was implemented on the computer.3 3. that used the random decrement technique. many identification algorithms have been developed to estimate frequency and damping from flight flutter data.1 Frequency.

and reduced flight test time when compared with previous methods of flight flutter testing. and (2) to develop effective noise reduction and transfer function enhancement (ref. . Reference 22 stated the need for completely automatic excitation.0 940485 Online. The subcritical damping trend for a 2. Future of Flight Flutter Testing The future of flight flutter testing has been defined at several times. A linear extrapolation of the data trend provided an instability airspeed prediction that agreed closely with the actual instability onset airspeed encountered. it is doubtful that sudden changes in damping. . more accurate. Two future needs identified from the papers presented at the symposium were (1) to further develop parameter estimation algorithms that would provide better estimates from noisy data and closely spaced modes (ref. Flight flutter testing was. The high-speed computer was identified as the tool for developing advanced data analysis methods that would more fully satisfy the desired objectives of flight flutter testing (ref.6 . 43 The future needs expressed in the 1958 symposium were partially met at the time of the 1975 symposium. 28).0-Hz mode indicated that as airspeed was increased. was. 15). although reference 47 warned that the current techniques may still not predict flutter for explosive flutter cases. can be predicted with the accuracy and timeliness required to avoid flutter. 1958 Flight Flutter Test Symposium The final sentence in reference 4. 15). Decay traces were obtained by using the flaperons to excite the structure with frequency dwells–quick stops.16 .6-Hz and 3.” Reference 17 stated that improvements needed were to shorten the time required to obtain data and to provide complete and higher quality data.2 . The damping was estimated using FFT algorithms. This set of data provided a unique opportunity to validate the accuracy of this algorithm because this configuration could be safely flown to a condition of zero damping. 46)..08 g Extrapolated Actual instability . 1975) techniques were faster and. which may occur between flutter test points. these two modes coupled and caused a decrease in the damping level. Flight of the airplane with this store configuration is characterized by an LCO. and the data were complete and of higher quality. The flutter testing symposia held in 1958 and 1975 identified future directions and needs (refs. Figure 11 shows the damping trend obtained in nearreal time for an F-16 airplane configured with AIM-9J missiles.8 Mach number Recommendations for Future Research and Development 1. In this instance. 42). real-time monitoring of aeroelastic stability during flight testing needs to be developed and implemented.04 0 .e. Although today’s techniques appear adequate to warn of the on-set of flutter for gradual decreases in damping. and 370-gal external fuel tanks (ref.4 . 45). The damping level decreased with increasing airspeed until it was no longer safe to continue the test. The time required to acquire the data was not significantly less because most of the flight test organizations were conducting sine sweeps at stabilized test points. increased safety. 1975 Flutter Testing Techniques Symposium The 1975 symposium contained several papers describing the application of techniques that used computers to estimate frequency and damping from flight flutter test data. Our future needs will then be presented.12 Damping. data-recording and data-reduction systems. 14 . Damping trend for a limited amplitude instability. more automated. The aircraft structure was excited by pilot-induced control surface pulses and damping was estimated using an FFT algorithm (ref. Testing was less hazardous in 1975 than in the 1950’s. Figure 11. which was presented at the 1958 flight flutter test symposium. and 44). GBU-8 stores.configured with winglets (ref. “It is hoped that improvements in test techniques will eventually result in flight flutter tests that will give all the information wanted and will be considerably less hazardous than they are today. Most agreed that the current (i. The way to fulfill this need was to automate the data-reduction equipment. the techniques used provided a sufficient and adequate warning of the onset of flutter mainly because the decrease in damping was gradual. These symposia proceedings will be reviewed to confirm the progress made toward those needs. Reference 48 recommended that frequency and damping estimates for clearance to the next flutter test point be made between flights.

The ideal display would show predicted frequency and damping values being compared with flight test values in real time. 16Goldman.” Aeroelastic and Structures Research Laboratory.L. M.” in “Proceedings. “Survey of Aircraft Subcritical Flight Flutter Testing Methods..D. 13Rosenbaum.. pp. III. New methods should be researched to permit a reliable determination of flutter speed at a speed that is well below the actual one. (Royal Aeronautical Society). August 1992. F. may be one approach... U. 12–15.” NASA SP-385. 12–20. Goldman. S. and Control Conference.W. R. and Abrams. and Piperias. Navy.” NASA SP-385. Volume V. 8Champion. 87-0909. vol.” August 1990. Navigation. R. Hilton Head.A. 6NACA Subcommittee on Vibration and Flutter.. Massachusetts Institute of Technology. Aeroplane.. Airplane. F. pp..H. “The Question of Spontaneous Wing Oscillations (Determination of Critical Velocity Through Flight-Oscillation Tests).. 159–166. R..” September 1990. “F-16C/D Block 40 With Advanced Medium-Range Air-to-Air Missile (AMRAAM) Flutter Flight Test Evaluation.C. Aerodynamic Flutter.. Rider.. 141–167. April 1987. October 1956.Techniques such as modal filtering (refs. L. The whole process of flight flutter testing needs to be fully automated so flight flutter testing can be done much faster but more safely.M. 121–125.. W. The technique proposed by Nissim (ref.W.” NACA RM-56I12.” in “Proceedings of the 34th Society of Experimental Test Pilots Symposium. 2Lancaster... no. no. Aerospace. 91–96. Garrick. and Wrisley.. pp.. “Limit Cycle Oscillation and Flight Flutter Testing. Such monitoring also eliminates the need for stabilized test points. 2.E. “Aircraft Flight Flutter Testing at the 3Collar. 11Borst. and Roha.. “Flight Flutter Testing the B-58 I. These responses could then be accurately analyzed with less sophisticated algorithms that are more able to run in real time. “F-117A Flight Test Program. 11.. 5. 10Farley. which is the acceleration from one test point to the next. D.T. T. “Torsional Vibrations of the Tail of an “Flight Flutter Testing of the P6M.” NASA Ames-Dryden Flight Research Facility..” NASA TM-100417. A. 9Norton. Melbourne. vol. P. M. 18.” in “Proceedings of the 1958 Flight Flutter Testing Symposium. March 1969. 1958. R. South Carolina. Noll.. July 1916. and Strome.A. pp. May 1988. 897–912.. 7French. “Final Report on the Investigation of Flight Flutter Testing Techniques for the Bureau of Aeronautics. 21st Annual Symposium. R.” Reports and Memoranda. Investigations on a Transavia PL12/T-400 Aircraft. 276. pp. no.” NASA CR-132479. B. L. February 1978. Real-time monitoring of stability eliminates the most hazardous part of flight flutter testing.G. 15Kehoe. D. August 1974.J. Australia.” AFFTC TR-92-19. REFERENCES 1Von Schlippe. 3. W. A. Broad-band excitation techniques also need to be developed so that a response signal of sufficient amplitude over the entire frequency range of interest is continually provided for real-time analysis. 1958. 17Mahaffey. pp. D. and Cabrera. December 1950.” I. E. 14Kachadourian. AIAA Guidance.” Aeronautical Research Laboratories. “E-6 Flutter Investigation and Experience. and Reed.. Jr.R. July 1989. E. H. which is extremely time consuming. M. December 1992. Society of Flight Test Engineers..4-12. 51).. Cooley. Moore. “Flutter Investigations Involving a Free Floating Aileron. pp. November 1981.” AIAA-92-4601-CP...W. Aircraft. pp. “History of Flight Flutter Testing. ed.” in “Proceedings of the 1958 Flight Flutter Testing Symposium. P.D. 15 . “A Survey and Evaluation of Flutter Research and Engineering... Testing 5Garrick. 49 and 50) can uncouple response measurements to produce simplified. “Historical Development of Aircraft Flutter.S. G.. and Zapata. C.” AIAA Paper no. “The First Fifty Years of Aeroelasticity.L.. 12Schwartz.4-1–3. J.” in “Proceedings of the 1958 Flight Flutter Symposium.” NACA TM-806. R. October 1936 (translation).” AIAA 81-0491.. which is based on identifying the coefficients of the equations of motion followed by solving of these equations to determine the flutter speed. “Flutter 4Tolve.” NASA SP-385. Aircraft Structures Technical Memorandum 515. single-degree-of-freedom responses. in “AIAA Selected Reprint Series. 1958.

“Flight Flutter 29Nordwall.” in “Proceedings of the 1958 Flight Flutter Testing Symposium.G. R.” in “Flutter Testing Techniques: A conference held at Dryden Flight Research Center.R.” NASA SP-385. 31Reed. 35Rhea. “Mobile Communications to Capture Testing Using Pulse Techniques.. Texas..” NASA CR-3152. R. Brosnan.. “The Evolution of Flutter Excitation at “Stabilizer Flutter Investigated by Flight Test.” in “Proceedings of the 4th AIAA Flight Test Conference. C. D.” NASA SP-385. H. September 19–22.” in “Proceedings of the 1958 Flight Flutter Testing Symposium.” in “Proceedings of the 1958 Flight Flutter Testing Symposium.B. pp.. Air Forces: Equipment and Technique. 4.K. 26Vernon. 19Meany. G.J. and Maxwell. M. pp. 24Bartley.” NASA SP-385. and Chen.M. and Lenk. H. 1974. R. Consumer Market.... W. 1958.D.R. M. 127–132.K. 34th AIAA Structures... and Brown. 22Dublin. C.” ONERA TP no.6-11. W. Ohio.. “A New Flight Flutter Excitation Demonstrator Flight Test Program Overview.H..C.” AIAA-83-2747. “Douglas Experience in Flight Flutter Testing. E. “Pyrotechnic Bonkers for Structural Tests in Flight. “X-31A Flight Flutter Test Excitation by Control Surfaces.H.1–V-1.” NASA SP-415. E.” NASA SP-385.” NASA SP385. V-1. W.” NASA SP-385. April 1993.” in “Proceedings. R. 34Meyer.. 73–81.. 39Allemang.H.B.J.” AIAA-93-1538-CP. 23Dobbs.. pp. California. September 18–21. Millet.” in “Proceedings of the 1958 Flight Flutter Testing Symposium. “X-29A Technology 25Reed. pp. 103–110. J. pp. V. August 1983. III. 1958. and Wittman... “Flight Flutter Testing of Multi-Jet Aircraft. B.J. F. La Jolla.. Foppe. Structural Dynamics and Materials Conference. “Flight Flutter Testing of Supersonic Interceptors.B. “Random Air Turbulence as a Flutter 38Katz.H.. Society of Flight Test Engineers. August 14–18.. McDonnell Aircraft.. and Peller.M. 1958. 32Philbrick. 1976. 69–72. Nevada. June 1979.” Aviation Week & Space Technology.” in “14th Annual Symposium Proceedings. New York.. December 1987.” in “Proceedings of the 1958 Flight Flutter Testing Symposium. 13th Annual Symposium. III. 16 . Modal Analysis and Dynamic Component Synthesis.7.. pp... and Grossman.. 1988. 1989. D.” AFWAL-TR-87-3069. 21Laure. P. 433–456. J. R.K. Flight Dynamics Laboratory.T..B. pp. 91–ff.” in “Flight Testing Technology: A State-of-the-Art Review. H. Arlington. C. 1958. System. 1975. 1389 E.” Society of Flight Test Engineers.” NASA SP-385. pp.” NASA TM-86809. 1993. and Smith.18Stringham.. “In-Flight Investigation of a Rotating Cylinder-Based Structural Excitation System for Flutter Testing. Ness. and Schneider.. pp. pp. October 9–10..” in “Proceedings of the 1958 Flight Flutter Testing Symposium.” NASA TM-4512. 413–431. “Real-Time Pilot Guidance System for Improved Flight Test Maneuvers. 1993. and Moore.T. pp. 36Sefic.. 37Perangelo.6-1–4. Wright-Patterson Air Force Base. I.J. 33Laidlaw.L. May 1986. and Piazzoli.” Reno. “F-15 Test Excitation Source. Sinder. E. Johnson..H. 111–120.” Society of Flight Test Engineers. pp. pp. P.. D.” in “Proceedings of the 1958 Flight Flutter Testing Symposium. 41–50. 1958. May 31.J. J.. Jr.. W.. and Hodson.. California 1988. November 1983. J.. “Experimental L.J. vol.J. 27Norton.. M..” NASA SP-415. and Beals. A. 1982. R.F.” San Diego. “The Application of Pulse Excitation to Ground and Flight Vibration Tests. Jr. L. 41–42.. 1976. M. 1958. S.. “YF-16 Flight Flutter Test Procedures.. 133–142. “A Flight Investigation of Oscillating 20Hodson. 30Baird. 20th Annual Symposium.L. Flight Flutter Test Program. “Determination of Subcritical Frequency and Damping From B-1 Flight Flutter Test Data. W.R. “Development of a Mobile Research Flight Test Support Capability. W. S. Dobbs.” in “Flutter Testing Techniques: A conference held at Dryden Flight Research Center.. “Flight Flutter Test Methodology at Grumman. and Waisanen. 28Brignac.” in “19th Annual Symposium Proceedings.. 1958..L.” Society of Flight Test Engineers.J.

” San Diego.. S. Manchester U.” NASA SP-415.” in “Proceedings of the 11th International Modal Analysis Conference. pp.. NASA SP-385. “Time Series Analysis in Flight Flutter Testing at the Air Force Flight Test Center: Concepts and Results. E. Program. and Gilyard.C.J.. Near-Real-Time Processing of Flutter Test Data. 75-1031.J. and Moody. Lecture Series 1992-01.. W.” in “Flutter Testing Techniques: A conference held at Dryden Flight Research Center.R. University of Manchester..” NASA SP-415. “Development of an On-line Modal State Monitor. Aeronautical Report no. “Some Experience Using Subcritical Response Methods in Wind-Tunnel Flutter Model Studies. “Introduction to Flutter of Winged 42Kehoe. J. 1976. R. 43NASA Langley Research Center.... R.” von Karman Institute for Fluid Dynamics.” Kissimmee.E. Proceedings of the “Development of an On-line Parameter Estimation System Using the Discrete Modal Filter. Katz. B.. 41Nash. Freudinger. 47Foughner. 1976. 17 . “KC-135 Winglet Flight Flutter Test Aircraft. 45Lenz. 377–394.. 1989. S. October 1975.W..C.R.. 51Nissim. pp. L. June 1981. California. D. California.” in “Proceedings of the 10th International Modal Analysis Conference. 1958 Flight Flutter Testing Symposium. Los Angeles. May 1958. pp. Florida. and Allemang. Flutter 50Shelley. February 1–4. Freudinger. 9105. Edwards Air Force Base.. 49Shelley. G. “Modal Parameter Identification Techniques for Flight Flutter Testing.J. H.B. 48Wright... 1993. 1992.W. “The Application of Digital Computers to J.” NASA SP-415.K.” NASA TP-2923. Jr.. and Allemang. 1976.. Air Force Flight Test Center. and McKeever. 173–183. California.J.T. L. February 3–7. Testing Techniques. “Method for Experimental Determination of Flutter Speed by Parameter Identification. M. pp... 181–191. 46Hurley. 1991. S..” AFFTC TR-81-4. NASA SP-415.E. December 1991.C. “F-15 Flight Flutter Testing: Aircraft Systems and Test Operations. J. 44National Aeronautics and Space Administration. 1975.” AIAA Paper no.” in “Flutter Testing Techniques: A conference held at Dryden Flight Research Center.” in “Flutter Testing Techniques: A conference held at Dryden Flight Research Center.40Cooper. 287–317.” in Aeronautical Engineering Group. AIAA 1975 Aircraft Systems and Technology Meeting. R.

DC 20546-0001 NASA TM-4720 11. 800 Elkridge Landing Road. 12a. digital data preprocessing. Kehoe 7. to Washington Headquarters Services. Suite 1204. SPONSORING/MONOTORING AGENCY NAME(S) AND ADDRESS(ES) 10. NUMBER OF PAGES Aeroelasticity. Rotterdam. including suggestions for reducing this burden. PERFORMING ORGANIZATION REPORT NUMBER NASA Dryden Flight Research Center P. Flutter test procedures 17. Paperwork Reduction Project (0704-0188). AGENCY USE ONLY (Leave blank) 2. 14.O. comments regarding the direction of future developments and needs are presented. Directorate for Information Operations and Reports. instrumentation systems. Arlington. DC 20503. LIMITATION OF ABSTRACT Unlimited Standard Form 298 (Rev. SUPPLEMENTARY NOTES Previously presented at the AGARD Structures and Materials Panel Meeting. SUBJECT TERMS 15. Finally. integrated effectiveness of the various approaches used. Flight flutter testing. DISTRIBUTION/AVAILABILITY STATEMENT 12b. SECURITY CLASSIFICATION OF ABSTRACT Unclassified Available from the NASA Center for AeroSpace Information. AUTHOR(S) Michael W. DISTRIBUTION CODE Unclassified—Unlimited Subject Category 05 13. May 8–10. California 93523-0273 H-2077 9. Send comments regarding this burden estimate or any other aspect of this collection of information. REPORT TYPE AND DATES COVERED Technical Memorandum 4. REPORT DATE October 1995 3. VA 22202-4302. and to the Office of Management and Budget. Flutter test history. SECURITY CLASSIFICATION OF REPORT Unclassified NSN 7540-01-280-5500 18. SPONSORING/MONITORING AGENCY REPORT NUMBER National Aeronautics and Space Administration Washington. 1. searching existing data sources. 1215 Jefferson Davis Highway. including the time for reviewing instructions. FUNDING NUMBERS A Historical Overview of Flight Flutter Testing WU 505-63-50 6. 2-89) Prescribed by ANSI Std. TITLE AND SUBTITLE 5. 1995. MD 21090. gathering and maintaining the data needed. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. Practical experiences and example test programs illustrate the combined.Form Approved OMB No. Linthicum Heights. Box 273 Edwards. Structural excitation systems. SECURITY CLASSIFICATION OF THIS PAGE Unclassified 19. PRICE CODE AO3 20. ABSTRACT (Maximum 200 words) This paper reviews the test techniques developed over the last several decades for flight flutter testing of aircraft. and completing and reviewing the collection of information. Washington. Z39-18 298-102 . and parameter identification algorithms (for frequency and damping estimates from the response data) are described. 0704-0188 REPORT DOCUMENTATION PAGE Public reporting burden for this collection of information is estimated to average 1 hour per response. Aircraft vibration. Netherlands. (301)621-0390 19 16.